Europe can finance a factory it cannot staff

There is more money available for European solar manufacturing than there has been at any point in twenty years. There are subsidy envelopes, resilience mandates, sovereign funds with a mandate to be patient, and a procurement environment that has started writing origin into the tender rather than the press release. On paper the constraint has been released.

It has not been released. It has moved.

What a plant actually consumes

A cell plant consumes three things in descending order of difficulty: a qualified process team, a power contract long enough to outlive the equipment, and money. The public conversation is arranged in exactly the reverse order, which tells you who has been doing the talking.

The team is the hard one because it cannot be procured. There is no market for four hundred people who have commissioned a line before, because the continent that had them stopped building lines around 2012 and those people went and did something else. You can hire the equipment vendor's commissioning engineers, and everyone does, and they leave — taking with them the specific knowledge of which of your machines lies about its throughput.

A subsidy can shorten a payback. It cannot shorten a learning curve.

The learning curve is the part nobody underwrites. First-year yield on a new line is not a number you negotiate; it is a number you discover, and the distance between a discovered 82% and a modelled 94% is the whole of the business case.

The power contract nobody wants to sign

The second constraint is more tractable and more boring. Cell manufacturing is electricity-intensive in a way that makes the electricity price a strategic input rather than a line item, and it needs that price fixed over a horizon longer than most utilities want to write.

That is a solvable problem — it is a contract, not a physics problem — but it is solved in a room where the manufacturer has almost no leverage, because the plant is a single site with a single load and no alternative. The plants that get built will be the ones that solved this before the ribbon-cutting, not after.

Why the announcements keep outrunning the concrete

Put those two together and the pattern in the last three years explains itself. The announcement stage requires money, which exists. The commitment stage requires a power contract, which is hard. The operating stage requires a team, which does not exist yet and takes four years to make.

So a great deal of capital is standing at the first gate looking pleased with itself.

What would change it

Nothing dramatic. Three unglamorous things, in order:

• Train the process team before the building is finished, on somebody else's line, and accept that you are paying to educate people who might leave.
• Sign the power contract before the equipment order, not after — the sequence is the entire negotiating position.
• Build one line properly rather than four lines optimistically, and let the yield curve on the first one price the next three.

None of that is a technology roadmap and none of it makes a good chart. It is also, as far as anyone can tell, the only version of this that ends with a factory.